[Bug 23565] cloneNode IDL does not match reality

https://www.w3.org/Bugs/Public/show_bug.cgi?id=23565

--- Comment #23 from Boris Zbarsky <bzbarsky@mit.edu> ---
On the other hand, for the common case of cloning, which is deep clones, the
API would be _way_ better if it could be just cloneNode().

The only real issue here is that there is legacy content that depends on
cloneNode(undefined) doing a shallow clone.  While we need to support that
legacy content, I believe that the default behavior for no arguments passed
should be the sane one.  That's why the mailing list discussion initially
decided the default here should be "true" to start with.

Does the argument for consistency between undefined and not passed really trump
the web developer ergonomics argument here?

> and will require some IDL hack long term to make it work.

That's putting spec purity ahead of developers, which is just wrong from a
priority of constituencies perspective...

-- 
You are receiving this mail because:
You are the QA Contact for the bug.

Received on Tuesday, 29 October 2013 22:11:17 UTC